[PATCH 0/2] iio: light: vcnl4000: Add support for Capella CM36686 and CM36672P

[PATCH 0/2] iio: light: vcnl4000: Add support for Capella CM36686 and CM36672P

[Erikas Bitovtas]

Capella CM36686 is a combined ambient light and proximity sensor
developed by Capella Microsystems, now a subsidiary of Vishay
Intertechnology Inc. It is fully compatible with Vishay VCNL4040 and its
driver can be used for Capella CM36686.

Capella CM36672P is a proximity sensor which is fully compatible with
CM36686, and in turn, with VCNL4040.

This patch series adds compatibles for cm36686 and cm36672p, as well as
disables the light channel for cm36672p.

Signed-off-by: Erikas Bitovtas <xerikasxx@gmail.com>
---
Erikas Bitovtas (2):
      dt-bindings: iio: light: vcnl4000: add Capella CM36686 and CM36672P
      iio: light: vcnl4000: add support for Capella CM36686 and CM36672P

 .../bindings/iio/light/vishay,vcnl4000.yaml        | 17 +++++----
 drivers/iio/light/vcnl4000.c                       | 40 ++++++++++++++++++++++
 2 files changed, 51 insertions(+), 6 deletions(-)
---
base-commit: 132737e360b4c0daa7f473faf0f55cb04ee3e15c
change-id: 20260210-cm36686-4d328b87b6d2

Best regards,
-- 
Erikas Bitovtas <xerikasxx@gmail.com>

[PATCH 1/2] dt-bindings: iio: light: vcnl4000: add Capella CM36686 and CM36672P

[Erikas Bitovtas]

Capella CM36686 is an ambient light and proximity sensor developed by
Capella Microsystems, now a subsidiary of Vishay Intertechnology Inc. It
has an I2C address of 0x60 and is fully compatible with an existing
driver for VCNL4040. Capella CM36672P is a proximity-only sensor that
is fully compatible with CM36686, and therefore with VCNL4040. Add
compatibles for cm36672p and cm36686, with a fallback for cm36686 of
vcnl4040.

Signed-off-by: Erikas Bitovtas <xerikasxx@gmail.com>
---
 .../devicetree/bindings/iio/light/vishay,vcnl4000.yaml  | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)

diff --git a/Documentation/devicetree/bindings/iio/light/vishay,vcnl4000.yaml b/Documentation/devicetree/bindings/iio/light/vishay,vcnl4000.yaml
index 4d1a225e8868..2ba4d5de4ec4 100644
--- a/Documentation/devicetree/bindings/iio/light/vishay,vcnl4000.yaml
+++ b/Documentation/devicetree/bindings/iio/light/vishay,vcnl4000.yaml
@@ -18,12 +18,17 @@ allOf:
 
 properties:
   compatible:
-    enum:
-      - vishay,vcnl4000
-      - vishay,vcnl4010
-      - vishay,vcnl4020
-      - vishay,vcnl4040
-      - vishay,vcnl4200
+    oneOf:
+      - enum:
+          - capella,cm36672p
+          - vishay,vcnl4000
+          - vishay,vcnl4010
+          - vishay,vcnl4020
+          - vishay,vcnl4040
+          - vishay,vcnl4200
+      - items:
+          - const: capella,cm36686
+          - const: vishay,vcnl4040
 
   interrupts:
     maxItems: 1

-- 
2.53.0

Re: [PATCH 1/2] dt-bindings: iio: light: vcnl4000: add Capella CM36686 and CM36672P

[Krzysztof Kozlowski]

On Tue, Feb 10, 2026 at 10:46:10PM +0200, Erikas Bitovtas wrote:
> Capella CM36686 is an ambient light and proximity sensor developed by
> Capella Microsystems, now a subsidiary of Vishay Intertechnology Inc. It
> has an I2C address of 0x60 and is fully compatible with an existing
> driver for VCNL4040. Capella CM36672P is a proximity-only sensor that
> is fully compatible with CM36686, and therefore with VCNL4040. Add
> compatibles for cm36672p and cm36686, with a fallback for cm36686 of
> vcnl4040.

There was CM36686 v1 and v2, so I do not understand versioning here.
Also, lack of any changelog.

And how is it supposed to work for us? Try yourself:

  $ b4 diff 20260210-cm36686-v1-1-aef68dd46ad4@gmail.com
  Could not find lower series to compare against.

Please implement previous feedback.

Best regards,
Krzysztof

Re: [PATCH 1/2] dt-bindings: iio: light: vcnl4000: add Capella CM36686 and CM36672P

[Erikas Bitovtas]

On 2/11/26 8:49 AM, Krzysztof Kozlowski wrote:
> 
> There was CM36686 v1 and v2, so I do not understand versioning here.
> Also, lack of any changelog.
> 
> And how is it supposed to work for us? Try yourself:
> 
>   $ b4 diff 20260210-cm36686-v1-1-aef68dd46ad4@gmail.com
>   Could not find lower series to compare against.
> 
> Please implement previous feedback.
> 
> Best regards,
> Krzysztof
> 

Since I am making changes to the existing driver instead of creating a new one,
I introduced a new patch series. As I mentioned in the cover letter, cm36686 is
fully compatible with vcnl4040, so instead of creating a new binding, I create a
fallback compatible for the device. I probably should have named this patch
series something else.
As for the previous feedback, I added the subject prefix and removed the
redundant phrasing.

Re: [PATCH 1/2] dt-bindings: iio: light: vcnl4000: add Capella CM36686 and CM36672P

[Krzysztof Kozlowski]

On 11/02/2026 11:07, Erikas Bitovtas wrote:
> 
> 
> On 2/11/26 8:49 AM, Krzysztof Kozlowski wrote:
>>
>> There was CM36686 v1 and v2, so I do not understand versioning here.
>> Also, lack of any changelog.
>>
>> And how is it supposed to work for us? Try yourself:
>>
>>   $ b4 diff 20260210-cm36686-v1-1-aef68dd46ad4@gmail.com
>>   Could not find lower series to compare against.

Look here.

>>
>> Please implement previous feedback.
>>
>> Best regards,
>> Krzysztof
>>
> 
> Since I am making changes to the existing driver instead of creating a new one,
> I introduced a new patch series. As I mentioned in the cover letter, cm36686 is
> fully compatible with vcnl4040, so instead of creating a new binding, I create a
> fallback compatible for the device. I probably should have named this patch
> series something else.

That's fine, but that's v3 of previous patches. Your work was to add
CM36686 support. How you do it, evolves, but patchset/work is one
continuous work. When you rework approach next time, you also start from
v1? And then you go back to previous solution of new driver it will jump
from v1 to v3?

> As for the previous feedback, I added the subject prefix and removed the
> redundant phrasing.

And how do I know it if there is no changelog with list of changes? Look
how many patches is here:
https://lore.kernel.org/linux-devicetree/

Best regards,
Krzysztof

Re: [PATCH 1/2] dt-bindings: iio: light: vcnl4000: add Capella CM36686 and CM36672P

[Erikas Bitovtas]

>>>
>>> Please implement previous feedback.
>>>
>>> Best regards,
>>> Krzysztof
>>>
>>
>> Since I am making changes to the existing driver instead of creating a new one,
>> I introduced a new patch series. As I mentioned in the cover letter, cm36686 is
>> fully compatible with vcnl4040, so instead of creating a new binding, I create a
>> fallback compatible for the device. I probably should have named this patch
>> series something else.
> 
> That's fine, but that's v3 of previous patches. Your work was to add
> CM36686 support. How you do it, evolves, but patchset/work is one
> continuous work. When you rework approach next time, you also start from
> v1? And then you go back to previous solution of new driver it will jump
> from v1 to v3?
> 

There has been a misunderstanding. I assumed that since I will no longer
be developing that driver, this warrants a new patch series. I apologize
for this.
Here is the changelog since v2:
- Remove the previous unnecessary proposed driver and bindings.
- Add a fallback compatible for cm36686 of vcnl4040.
- Add a new compatible for cm36672p.
- Add channel info for cm36672p.
- Remove redundant information in the dt-bindings commit message.
Here is the link to v2:
https://lore.kernel.org/linux-iio/20260209182432.00006c3c@huawei.com/

I have received some feedback regarding the changes I made to the
existing vcnl4000 driver. Shall I submit the implementation of it as a
v3 to that series of patches?

Re: [PATCH 1/2] dt-bindings: iio: light: vcnl4000: add Capella CM36686 and CM36672P

[Andy Shevchenko]

On Wed, Feb 11, 2026 at 03:32:41PM +0200, Erikas Bitovtas wrote:

> >>> Please implement previous feedback.
> >>
> >> Since I am making changes to the existing driver instead of creating a new one,
> >> I introduced a new patch series. As I mentioned in the cover letter, cm36686 is
> >> fully compatible with vcnl4040, so instead of creating a new binding, I create a
> >> fallback compatible for the device. I probably should have named this patch
> >> series something else.
> > 
> > That's fine, but that's v3 of previous patches. Your work was to add
> > CM36686 support. How you do it, evolves, but patchset/work is one
> > continuous work. When you rework approach next time, you also start from
> > v1? And then you go back to previous solution of new driver it will jump
> > from v1 to v3?
> > 
> 
> There has been a misunderstanding. I assumed that since I will no longer
> be developing that driver, this warrants a new patch series. I apologize
> for this.
> Here is the changelog since v2:
> - Remove the previous unnecessary proposed driver and bindings.
> - Add a fallback compatible for cm36686 of vcnl4040.
> - Add a new compatible for cm36672p.
> - Add channel info for cm36672p.
> - Remove redundant information in the dt-bindings commit message.
> Here is the link to v2:
> https://lore.kernel.org/linux-iio/20260209182432.00006c3c@huawei.com/
> 
> I have received some feedback regarding the changes I made to the
> existing vcnl4000 driver. Shall I submit the implementation of it as a
> v3 to that series of patches?

I guess v4 would be better as this is assumed misversioned v3, if I got
the situation correctly.

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH 1/2] dt-bindings: iio: light: vcnl4000: add Capella CM36686 and CM36672P

[Erikas Bitovtas]

On 2/11/26 3:37 PM, Andy Shevchenko wrote:
> On Wed, Feb 11, 2026 at 03:32:41PM +0200, Erikas Bitovtas wrote:
> 
>>>>> Please implement previous feedback.
>>>>
>>>> Since I am making changes to the existing driver instead of creating a new one,
>>>> I introduced a new patch series. As I mentioned in the cover letter, cm36686 is
>>>> fully compatible with vcnl4040, so instead of creating a new binding, I create a
>>>> fallback compatible for the device. I probably should have named this patch
>>>> series something else.
>>>
>>> That's fine, but that's v3 of previous patches. Your work was to add
>>> CM36686 support. How you do it, evolves, but patchset/work is one
>>> continuous work. When you rework approach next time, you also start from
>>> v1? And then you go back to previous solution of new driver it will jump
>>> from v1 to v3?
>>>
>>
>> There has been a misunderstanding. I assumed that since I will no longer
>> be developing that driver, this warrants a new patch series. I apologize
>> for this.
>> Here is the changelog since v2:
>> - Remove the previous unnecessary proposed driver and bindings.
>> - Add a fallback compatible for cm36686 of vcnl4040.
>> - Add a new compatible for cm36672p.
>> - Add channel info for cm36672p.
>> - Remove redundant information in the dt-bindings commit message.
>> Here is the link to v2:
>> https://lore.kernel.org/linux-iio/20260209182432.00006c3c@huawei.com/
>>
>> I have received some feedback regarding the changes I made to the
>> existing vcnl4000 driver. Shall I submit the implementation of it as a
>> v3 to that series of patches?
> 
> I guess v4 would be better as this is assumed misversioned v3, if I got
> the situation correctly.
> 

Yes, I started a new patch series when I should have continued previous
one. I will submit an implementation of your feedback as a v4 to the
patch series I started initially. Thank you!

Re: [PATCH 1/2] dt-bindings: iio: light: vcnl4000: add Capella CM36686 and CM36672P

[Andy Shevchenko]

On Wed, Feb 11, 2026 at 03:44:37PM +0200, Erikas Bitovtas wrote:
> On 2/11/26 3:37 PM, Andy Shevchenko wrote:
> > On Wed, Feb 11, 2026 at 03:32:41PM +0200, Erikas Bitovtas wrote:

...

> >>>>> Please implement previous feedback.
> >>>>
> >>>> Since I am making changes to the existing driver instead of creating a new one,
> >>>> I introduced a new patch series. As I mentioned in the cover letter, cm36686 is
> >>>> fully compatible with vcnl4040, so instead of creating a new binding, I create a
> >>>> fallback compatible for the device. I probably should have named this patch
> >>>> series something else.
> >>>
> >>> That's fine, but that's v3 of previous patches. Your work was to add
> >>> CM36686 support. How you do it, evolves, but patchset/work is one
> >>> continuous work. When you rework approach next time, you also start from
> >>> v1? And then you go back to previous solution of new driver it will jump
> >>> from v1 to v3?
> >>>
> >>
> >> There has been a misunderstanding. I assumed that since I will no longer
> >> be developing that driver, this warrants a new patch series. I apologize
> >> for this.
> >> Here is the changelog since v2:
> >> - Remove the previous unnecessary proposed driver and bindings.
> >> - Add a fallback compatible for cm36686 of vcnl4040.
> >> - Add a new compatible for cm36672p.
> >> - Add channel info for cm36672p.
> >> - Remove redundant information in the dt-bindings commit message.
> >> Here is the link to v2:
> >> https://lore.kernel.org/linux-iio/20260209182432.00006c3c@huawei.com/
> >>
> >> I have received some feedback regarding the changes I made to the
> >> existing vcnl4000 driver. Shall I submit the implementation of it as a
> >> v3 to that series of patches?
> > 
> > I guess v4 would be better as this is assumed misversioned v3, if I got
> > the situation correctly.
> 
> Yes, I started a new patch series when I should have continued previous
> one. I will submit an implementation of your feedback as a v4 to the
> patch series I started initially. Thank you!

Do not forget to include the _full_ changelog per version

  ...from v<N>:
  ...
  ...from v2:
  ...
  ...from v1:
  ...

-- 
With Best Regards,
Andy Shevchenko

[PATCH 2/2] iio: light: vcnl4000: add support for Capella CM36686 and CM36672P

[Erikas Bitovtas]

Add support for Capella's CM36686 and CM36672P sensors. Capella
CM36686 is an ambient light and proximity sensor that is fully
compatible with VCNL4040 and can be used as is. For CM36672P, which is
a proximity-only sensor, also remove the IIO_LIGHT channel.

Signed-off-by: Erikas Bitovtas <xerikasxx@gmail.com>
---
 drivers/iio/light/vcnl4000.c | 40 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 40 insertions(+)

diff --git a/drivers/iio/light/vcnl4000.c b/drivers/iio/light/vcnl4000.c
index a36c23813679..97dd930d57f6 100644
--- a/drivers/iio/light/vcnl4000.c
+++ b/drivers/iio/light/vcnl4000.c
@@ -185,6 +185,7 @@ static const int vcnl4040_ps_oversampling_ratio[] = {1, 2, 4, 8};
 #define VCNL4000_SLEEP_DELAY_MS	2000 /* before we enter pm_runtime_suspend */
 
 enum vcnl4000_device_ids {
+	CM36672P,
 	VCNL4000,
 	VCNL4010,
 	VCNL4040,
@@ -240,6 +241,8 @@ static const struct i2c_device_id vcnl4000_id[] = {
 	{ "vcnl4020", VCNL4010 },
 	{ "vcnl4040", VCNL4040 },
 	{ "vcnl4200", VCNL4200 },
+	{ "cm36672p", CM36672P },
+	{ "cm36686", VCNL4040 },
 	{ }
 };
 MODULE_DEVICE_TABLE(i2c, vcnl4000_id);
@@ -1842,6 +1845,22 @@ static const struct iio_chan_spec vcnl4040_channels[] = {
 	}
 };
 
+static const struct iio_chan_spec cm36672p_channels[] = {
+	{
+		.type = IIO_PROXIMITY,
+		.info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |
+			BIT(IIO_CHAN_INFO_INT_TIME) |
+			BIT(IIO_CHAN_INFO_OVERSAMPLING_RATIO) |
+			BIT(IIO_CHAN_INFO_CALIBBIAS),
+		.info_mask_separate_available = BIT(IIO_CHAN_INFO_INT_TIME) |
+			BIT(IIO_CHAN_INFO_OVERSAMPLING_RATIO) |
+			BIT(IIO_CHAN_INFO_CALIBBIAS),
+		.ext_info = vcnl4000_ext_info,
+		.event_spec = vcnl4040_event_spec,
+		.num_event_specs = ARRAY_SIZE(vcnl4040_event_spec),
+	}
+};
+
 static const struct iio_info vcnl4000_info = {
 	.read_raw = vcnl4000_read_raw,
 };
@@ -1867,6 +1886,19 @@ static const struct iio_info vcnl4040_info = {
 };
 
 static const struct vcnl4000_chip_spec vcnl4000_chip_spec_cfg[] = {
+	[CM36672P] = {
+		.prod = "CM36672P",
+		.init = vcnl4200_init,
+		.measure_proximity = vcnl4200_measure_proximity,
+		.set_power_state = vcnl4200_set_power_state,
+		.channels = cm36672p_channels,
+		.num_channels = ARRAY_SIZE(cm36672p_channels),
+		.info = &vcnl4040_info,
+		.irq_thread = vcnl4040_irq_thread,
+		.int_reg = VCNL4040_INT_FLAGS,
+		.ps_it_times = &vcnl4040_ps_it_times,
+		.num_ps_it_times = ARRAY_SIZE(vcnl4040_ps_it_times),
+	},
 	[VCNL4000] = {
 		.prod = "VCNL4000",
 		.init = vcnl4000_init,
@@ -2053,6 +2085,14 @@ static const struct of_device_id vcnl_4000_of_match[] = {
 		.compatible = "vishay,vcnl4200",
 		.data = (void *)VCNL4200,
 	},
+	{
+		.compatible = "capella,cm36672p",
+		.data = (void *)CM36672P,
+	},
+	{
+		.compatible = "capella,cm36686",
+		.data = (void *)VCNL4040,
+	},
 	{ }
 };
 MODULE_DEVICE_TABLE(of, vcnl_4000_of_match);

-- 
2.53.0

Re: [PATCH 2/2] iio: light: vcnl4000: add support for Capella CM36686 and CM36672P

[Andy Shevchenko]

On Tue, Feb 10, 2026 at 10:46 PM Erikas Bitovtas <xerikasxx@gmail.com> wrote:
>
> Add support for Capella's CM36686 and CM36672P sensors. Capella
> CM36686 is an ambient light and proximity sensor that is fully
> compatible with VCNL4040 and can be used as is. For CM36672P, which is
> a proximity-only sensor, also remove the IIO_LIGHT channel.

A couple of nit-picks, otherwise LGTM (much better than the first
version!), thanks.

...

>         { "vcnl4020", VCNL4010 },
>         { "vcnl4040", VCNL4040 },
>         { "vcnl4200", VCNL4200 },
> +       { "cm36672p", CM36672P },
> +       { "cm36686", VCNL4040 },

Keep it ordered by device name (string literal).

...

> +static const struct iio_chan_spec cm36672p_channels[] = {
> +       {
> +               .type = IIO_PROXIMITY,
> +               .info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |
> +                       BIT(IIO_CHAN_INFO_INT_TIME) |
> +                       BIT(IIO_CHAN_INFO_OVERSAMPLING_RATIO) |
> +                       BIT(IIO_CHAN_INFO_CALIBBIAS),
> +               .info_mask_separate_available = BIT(IIO_CHAN_INFO_INT_TIME) |
> +                       BIT(IIO_CHAN_INFO_OVERSAMPLING_RATIO) |
> +                       BIT(IIO_CHAN_INFO_CALIBBIAS),
> +               .ext_info = vcnl4000_ext_info,
> +               .event_spec = vcnl4040_event_spec,
> +               .num_event_specs = ARRAY_SIZE(vcnl4040_event_spec),

> +       }

Keep trailing comma when it's not a terminator entry.

> +};

-- 
With Best Regards,
Andy Shevchenko